It must be pointless because you can't afford a nice 30" monitor
It would be silly of me to get 5970 or 580 just to play 1680z1050 when a $100 5770 is confident @ that resolution.
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
GeForce GTX 580 vs. Radeon HD 5970 2GB Performance @ [H]
- Thread starter FrgMstr
- Start date
It must be pointless because you can't afford a nice 30" monitor
It would be silly of me to get 5970 or 580 just to play 1680z1050 when a $100 5770 is confident @ that resolution.
I would bet that at 1920x1200 the differences between the two are far less pronounced.
Agreed. Although at the price point of the Radeon HD 5970 4GB I'd probably opt for the GeForce GTX 580 instead.
kcthebrewer
[H]ard|Gawd
- Joined
- Jun 14, 2004
- Messages
- 1,728
Agreed.
GoodBoy
2[H]4U
- Joined
- Nov 29, 2004
- Messages
- 2,760
You are missing the point.
The "subjective" gameplay (or the better feeling, smoother, consistent experience) can often differ from what a simple benchmark will show. This is precisely why [H] does their reviews differently and why we all come here instead of any one of the other cookie cutter review sites.
What this review shows is that the 5970 2Gb card is memory limited. It's also got the standard fare dual GPU issues. But someone just looking at a benchmark is going to think the 5970 is a better card for about the same money. Then they go to a lan party, their buddy has a gtx580, and WTF but his runs alot smoother! Mine should be winning! MINE!
But no. If you had a 4gb version of the 5970, the memory issue would likely be gone. Of course that card costs $1200 vs a $500 gtx580. Even SLI'd gtx580 is cheaper than that, which would offer performance well beyond a 4gb 5970. To be fair the $1200 version of the card is more aimed at ATI fans/collectors and is a limited run. But the performance/price on it is crap.
The main point the article is trying to make is that if you are in the market for a video card and want to spend about $500, with the available choices the single GPU solution is the better choice. I for one am very glad [H] is around doing these.
You are of course free to buy what ever you choose. If you feel these reviews are not honest, go to some other website for reviews. Frankly these reviews are the most honest ones out there.
PrincessFrosty
Supreme [H]ardness
- Joined
- May 6, 2009
- Messages
- 5,905
From what I've seen elsewhere and with the [H] review as well is that while the frame rate of the 5970 is more erratic (and that's to be expected with multi-GPU), it's frame rate lead is sufficient so that it's minimum frame rate is often just as good or better than a single GPU card. reviews like the anandtech confirm these findings across a much broader set of games.
Your parting shot about horrid drive support is telling, not only that you're eyeing up Nvidia cards already, but also that you're naive enough to think that Nvidia driver support is any better. Sure your current drive issues will go, and be replaced with an entirely new set of problems. Anyway, that's a tangent I wont follow any further.
Again this is tangential but fundamental I feel, these sorts of frame rates aren't really fantastic, especially for a hardware enthusiast site, if you were reviewing the 5770 or something for the run of the mill middle to low end PCs with not much rendering power to play with you might expect 30fps to be your baseline, but £500 video cards @ 2560x1600?
Personally as an enthusiast the target frame rate is too low, I've mentioned this before as a massive failing in the review structure. Hardocps methods are more specific than "canned benchmarks" which they downplay the usefulness of but it's a 0 or 1 result, if you don't agree with the subjective result of "playable" from the [H] team then the benchmarks actually don't tell you much at all, and canned benchmarks while less detailed offer better insight.
Furthermore the testing methodology doesn't take into account the benefit of the "remainder" frame rate when setting a game to 1 configuration such as 2560x1600 8xMSAA and both cards get playable frame rates, but one frame rate is quite a bit higher, they're seen as equal in that game but some of us see benefit in that additional frame rate, primarily because I don't agree with the target frame rates to begin with. Or put another way with a higher standard for optimal frame rate it might actually be that the difference is FPS between the cards is enough that at this new standard, one card actually does need to have it's settings lowerd where another doesn't. That's why canned benchmarks with graphed results from a wide range of settings are actually more helpful, at least to some of us.
The additional frame rate in something like Metro2033 is just as valuable to me as the lack of stuttering is to other people, which is why subjective reviews are a dangerous grey area. Certainly if i'm provided with overwhelming evidence that the average, as well as the min/max FPS numbers for a specific card are better, yet game play is deemed worse by subjective analysis, then there better at least be investigation into why, or some kind of detailed reason as to why I should just abandon the frame rate numbers and trust a subjective measurement.
I refuse to believe we cannot measure it, I think someone with a statistics background might be able to give a better method of measurement, maybe the standard deviation of the frame rate to measure a kind of jitter, the problem is this is often just as much engine based and different per game than all the other things, you have to shoot for an average over many games since you don't know what the reader will be playing...which is why i measure general/overal performance in wins/losses over a number of games and see the 5970 winning here, and at anandtech.
Actually you mean the gaming frame rate isn't as consistent in SOME games, the frame rate looks just as consistent to me in the last 3 games, the fame rate doesn't look any more stable for either card...with such a small game sample size we could say that its likely there games where the 5970 might appear more stable.
Last edited:
Tamlin_WSGF
2[H]4U
- Joined
- Aug 1, 2006
- Messages
- 3,114
I think that Brent meant that since the 580 have playable framerates as the 5970 in all games at same settings, while being able to have higher settings in some games with playable framerates, its a better card. Its the overall experience, not how many games it wins according to framerate.
The 5970 is faster and cheaper if you don't use 8X AA/2560x1600 or similar situations where the 1GB memory is limiting the card, so in those cases the 5970 is the better card. As you could see with 4X AA/2560x1600, it was 25% faster then the 580 while still being cheaper then the 580. But, if you have a 2560x1600 monitor, the 580 would still be the better choice, since you can use 8X AA on it.
You have experienced it so you know where those stutters and texture thrashing come from.
Let me ask you a question. Would you rather have a card that stutter or dip into teens or not? I know it's an obvious question but you seem to want a card that stutters just because it has better average frame rate.
It isn't just a single dip in F1. There are many dips. Just because it does hit magic 30fps doesn't mean it isn't thrashing all over the place. If your frames fluctuate you will see more jerking than a card that get constant 30-40 fps.
It isn't about what card wins @ what benchmark. It is about gaming experience that is why HARDOCP is writing these kind of articles.
Tamlin_WSGF
2[H]4U
- Joined
- Aug 1, 2006
- Messages
- 3,114
She's not dead wrong. Some might argue that you don't need more then 4X AA on a 2560x1600 resolution and that 25% higher framerate is preferable. In those cases, the 5970 would be the better card according to gameplay experience. Especially considering that the 5970 is cheaper. Its a matter of opinion. However, [H] preferred 8X AA 2560x1600 instead of higher framerates and on those preferences/conditions, I find [H] correct to say that the 580 is a better card.
Last edited:
kcthebrewer
[H]ard|Gawd
- Joined
- Jun 14, 2004
- Messages
- 1,728
PrincessFrosty, how is a ~30fps minimum framerate not playable? What minimum framerate do you think is fair to maintain 100% playability?
If 30FPS is maintained at all times as a minimum then yeah, I'd agree that it's enough. However 30FPS average isn't necessarily playable. You may end up with a slide show in large firefights. That's not acceptable.
At least nVidia's drivers didn't completely break my games. Two games that I love to play I cannot, because ATi still can't seem to figure out OpenGL. Sorry, but from using cards from both camps, I can say that NV's drivers are less buggy. Don't even get me started on the CCC. So yes, I am "eying up" a 580 right now. The only way I would not pull the trigger on a GTX 580 is if the 69xx's blow it clean out of the water. If performance is close though, the GTX 580 will still be looking good for me.
That's_Corporate
[H]ard|Gawd
- Joined
- Sep 30, 2006
- Messages
- 1,738
Sure, you'd sum the squared variances from the mean and divide by the total number of FPS measurements over the sample. But then all you get is a number, and that number means absolutely squat to 99% of people as it takes a pretty decent quantitative background to figure it out.
Not to mention, you'd need to phrase it within a probability distribution so you could graphically see the skewness or kurtosis which is present.
Then again, me, and about one other person might find that cool
The rest would say WTF.
addictedto60fps
[H]ard|Gawd
- Joined
- Mar 31, 2010
- Messages
- 1,329
Hahah, yeah, I would be one of the ones who are like wtf. This is pretty simple, if you ask me:
Generally more stable frame rates at 2560x1600 resolution with 8x MSAA, Physx, 3D, tesellation, single card solution, (arguably better drivers) = Nvidia
Mostly higher frame rates at 2560x1600 with 4x MSAA, cheaper, Eyefinity = ATI
So, whichever is more important to you, you pick that one.
Generally more stable frame rates at 2560x1600 resolution with 8x MSAA, Physx, 3D, tesellation, single card solution, (arguably better drivers) = Nvidia
Mostly higher frame rates at 2560x1600 with 4x MSAA, cheaper, Eyefinity = ATI
So, whichever is more important to you, you pick that one.
But they sure have broken computers....LOL. I couldn't resist.
If that is the case why stop @ 2560x1500 4xAA? Why do we need more than 1080p no AA? It's a useless argument.
Tamlin_WSGF
2[H]4U
- Joined
- Aug 1, 2006
- Messages
- 3,114
Its not a useless argument if you understand it (where its coming from).
First of all, you have the high resolution (more pixels making the same image) and low pixel pitch (smaller pixels) of the 30" 2560x1600 making aliasing less visible, secondly, going 1x MSAA to 4X MSAA is a big change, 4X AA to 8X AA is not that big. Personally, I prefer alternative AA methods to going 8X MSAA. Others might prefer 4X MSAA + 25% higher framerate(vs. 580 with 4X AA) to 8X MSAA. Its subjective and therefore its not wrong if she feels this way.
Secondly, you need more then 1X AA (what you call no AA) on 1080P, since it removes aliasing, and since aliasing is more visible on 1080P then 1600P you need more of it (more anti-aliasing) . However, people have used 1X AA in games like Crysis due to framerate being more important. Again, it depends on both situation and the subjective opinion of the gamer.
My point is that though I feel its correct of [H] to say that the 580 is the better card, since it can perform with playable framerate in all games and higher settings in some, doesn't exclude PrincessFrosty also being correct from her point of view. its a matter of preferences.
[H]'s review shows that the 580 is better to get with a 2560x1600 resolution running 8X AA due to memory bottleneck, but it also shows that the 5970 is better to get if you won't run into the bottleneck also. After all, it was the bottleneck in the review that degraded the 5970 gameplay experience. Besides the bottleneck the 5970 was faster then the 580 and cheaper.
Last edited:
I think that Brent meant that since the 580 have playable framerates as the 5970 in all games at same settings, while being able to have higher settings in some games with playable framerates, its a better card. Its the overall experience, not how many games it wins according to framerate.
The 5970 is faster and cheaper if you don't use 8X AA/2560x1600 or similar situations where the 1GB memory is limiting the card, so in those cases the 5970 is the better card. As you could see with 4X AA/2560x1600, it was 25% faster then the 580 while still being cheaper then the 580. But, if you have a 2560x1600 monitor, the 580 would still be the better choice, since you can use 8X AA on it.
What about the fact the 5970 can do Eyefinity on a single video card? Nvidia requires TWO for NV Surround, therefore double the outlay. Not that I'm complaning since I've got a pair of 480s myself and imo the extra performance is needed for triple wide resolutions, however that does affect the overall value equation imo.
Tamlin_WSGF
2[H]4U
- Joined
- Aug 1, 2006
- Messages
- 3,114
That is true, but the reviewer bases the review and conclusion upon the given conditions. Eyefinity and surround solutions is not a part of this review. For someone using 1920x1200 or similar resolution, the conclusion would probably change as well, since its not likely that the user will ever encounter the memory bottleneck.
Fact remains that the memory bottleneck the 5970 showed at 8X AA/2560x1600 gave the reviewer a worse gameplay experience even though framerate was higher. In all other cases where there were no bottleneck, the 5970 performed better then the 580, but the reviewer bases his conclusion on the overall gaming experience and since the 580 have no bottleneck in those cases, it was more consistent.
GoldenTiger
Fully [H]
- Joined
- Dec 2, 2004
- Messages
- 29,666
I've had a 27" 1920x1200 monitor for awhile, now....
Actually, this is a common misconception: 2560x1600 does NOT eliminate jaggies inherently, as the dot pitch is the same between pixels as is the size compared to, say, a 24" 1920x1200 panel. You do have a little bit less aliasing since any given shape has more pixels to be drawn with, however aliasing is very much visible and a large issue even at 30" 2560x1600. I have owned my Dell 3007wfp-hc for a little short of two years now. I still prefer to run at minimum 4x MSAA, and often 8x CSAA, in my games. It still is an immense image quality improvement at these high resolutions.
It's a useless argument because there's no set settings you should be playing your PC games. That's what consoles are for.,
I have smaller pixel pitch than 30" 2560x1600 monitor as I have 21.5" 1080P monitor.
30" 2560x1600 pixel pitch 0.250
21.5" 1920x1080 pixel pitch 0.248
I can definitely tell aliasing even in my smaller pixel pitched monitor. You keep mentioning matter of preference but you try to justify your preference and eye candy.
It's not like 5970 is slow @ 2560x1600 4xAA. If I could I rather turn that extra pixel pushing power into image quality improvements but a 5970 can't without performance issues. That is the point of the H's article.
I would think an enthusiast who's going to spend $3000 on a monitor and $500-$1000 on a video cards would want to turn up the settings without performance issues. Hey I could be wrong. These spenders are probably playing 640x480 so they can get 100 more fps.
Lord_Exodia
Supreme [H]ardness
- Joined
- Sep 29, 2005
- Messages
- 7,008
Some people a genuinely understand both scenarios(TamlinWSGF), other people want to defend their purchase of their 5970. Tamlin is saying that she may prefer to have the extra performance and run at 4x or below and that's fine, as some people can choose that and be fine, however I dont believe that is the truth behind it from her. I have a hard time believing someone who fails to acknowledge time and time again the many driver issues present on the 5970. If you ask princessfrosty if she ever experienced driver issues I've seen her say nothing negative about it. However ask someone who has no problem admitting the truth such as Mark Warner or another owner and they will tell you how horrible of a solution it can be some times. It's not an overall horrible solution but many times you need to disable crossfire and run a single 1gb 5870 speed card or roll drivers back and forward.
In the end I'll quote a poster from very recently who posted in a thread about his feelings of his 5970
Source
http://hardforum.com/showthread.php?p=1036423650#post1036423650
So I guess it does make sense that they want to fight for more FPS as much as they can and minimize IQ, as they are too often that not disabling crossfire and lowering IQ for playable settings on their single 1GB 5870 performing card.
Last edited:
Interesting article but at 2560x1600 do you really need 8xAA anyway? I mean the difference between 8xAA and 4xAA is far less pronounced that the difference between 2xAA and 4xAA to me even at my preferred gaming resolution of 1920x1200. So in that respect the HD 5970 is the faster card for gaming at that resolution provided you aren't bothered about having the absolute best image quality (and, if you are, then why would stop at 8xAA, why not 32xAA which is the maximum supported by the GTX 580?).
I've just switched from an HD 5870 CrossFireX setup to a single GTX 580 purely because I'm fed up of AMD's (IMO) disappointing drivers. If a game isn't scaling negatively or had CFX disabled completely (e.g. Blur, Darksiders, Dead Space, Dead Rising 2, etc.) then the scaling for the second GPU is usually disappointing. Then there's the fact that you have to wait weeks or months for AMD to support new games, which by the time they do means I've moved on to something else. It took them over FIVE months to fix an annoying slowdown bug with CFX in Prince of Persia: The Forgotten Sands for example (the framerate would drop to sub-30fps if the prince died and stay that way until you Alt-Tabbed to the desktop and back).
I've noticed that the games I'm playing on the GTX 580 feel a lot smoother than they did with HD 5870 CFX even though my average framerates are lower for many games (but not all of them surprisingly). Outside of benchmarks, this loss of performance isn't an issue because I play at 60 fps with v-sync and the GTX 580 is more than capable of reaching that for the majority of the games I've tested. The fact that it's a single card means there's less chance of microstuttering or framerate drops but due to the GTX 580 having 50% more video RAM than the 1 GB HD 5870 cards it means that it is less likely to run out of RAM which is usually the cause of hitching and sudden framerate drops.
Using MSI Afterburner's OSD to monitor GPU memory usage I noticed a lot of games exceeding 1 GB at a resolution of 1920x1200 (the native resolution of my 24" display) with high anti-aliasing. Add in triple buffering (I play my games with v-sync enabled) and that memory usage increases further. For example, Lost Planet 2 DX11 runs much smoother with 8xAA on my GTX 580 than it did on my HD 5870 CFX setup and that's reflected even in the benchmarks which have higher framerates. That game can use up to 1.3 GB of video memory hence the low framerates I was getting before on the ATI cards. Hell, the game is even playable with 16xQCSAA! Also I can max out GTA IV with 100% draw/detail distance and Very High shadows/night shadows and the game still hits 60 fps which certainly wasn't the case with my HD 5870s which stuttered and hitched as the game ran out of video memory. Crysis Warhead is another game that can use 1.2 GB of video memory and remains completely playable on Enthusiast settings even at 16xAA!!!
Overall, I'm very impressed with this card and it makes a great replacement for CFX offering me smoother gameplay and (hopefully) fewer issues with future games due to the fact that it is no longer as dependent on regular driver updates like multi-GPU gaming is.
I've just switched from an HD 5870 CrossFireX setup to a single GTX 580 purely because I'm fed up of AMD's (IMO) disappointing drivers. If a game isn't scaling negatively or had CFX disabled completely (e.g. Blur, Darksiders, Dead Space, Dead Rising 2, etc.) then the scaling for the second GPU is usually disappointing. Then there's the fact that you have to wait weeks or months for AMD to support new games, which by the time they do means I've moved on to something else. It took them over FIVE months to fix an annoying slowdown bug with CFX in Prince of Persia: The Forgotten Sands for example (the framerate would drop to sub-30fps if the prince died and stay that way until you Alt-Tabbed to the desktop and back).
I've noticed that the games I'm playing on the GTX 580 feel a lot smoother than they did with HD 5870 CFX even though my average framerates are lower for many games (but not all of them surprisingly). Outside of benchmarks, this loss of performance isn't an issue because I play at 60 fps with v-sync and the GTX 580 is more than capable of reaching that for the majority of the games I've tested. The fact that it's a single card means there's less chance of microstuttering or framerate drops but due to the GTX 580 having 50% more video RAM than the 1 GB HD 5870 cards it means that it is less likely to run out of RAM which is usually the cause of hitching and sudden framerate drops.
Using MSI Afterburner's OSD to monitor GPU memory usage I noticed a lot of games exceeding 1 GB at a resolution of 1920x1200 (the native resolution of my 24" display) with high anti-aliasing. Add in triple buffering (I play my games with v-sync enabled) and that memory usage increases further. For example, Lost Planet 2 DX11 runs much smoother with 8xAA on my GTX 580 than it did on my HD 5870 CFX setup and that's reflected even in the benchmarks which have higher framerates. That game can use up to 1.3 GB of video memory hence the low framerates I was getting before on the ATI cards. Hell, the game is even playable with 16xQCSAA! Also I can max out GTA IV with 100% draw/detail distance and Very High shadows/night shadows and the game still hits 60 fps which certainly wasn't the case with my HD 5870s which stuttered and hitched as the game ran out of video memory. Crysis Warhead is another game that can use 1.2 GB of video memory and remains completely playable on Enthusiast settings even at 16xAA!!!
Overall, I'm very impressed with this card and it makes a great replacement for CFX offering me smoother gameplay and (hopefully) fewer issues with future games due to the fact that it is no longer as dependent on regular driver updates like multi-GPU gaming is.
Last edited:
hah, I ran beta CS on 15 frames. Was awesome and surprisingly "smooth."
I've been PC gaming for more than 15 years and I've never considered that good performance.
15 years oooooo oldtimer!
Like I said...spoilt!
Nothing spoiled about it. Up until about 10 years ago or so I wasn't able to afford good hardware. I suffered playing on crappy machines. I hated it but it was the best I could afford.
15 years? Im sure there's quite a few been playing significantly longer than that. I remember playing video games even before windows 3.1. So that would put it more than 17 (in december) years ago. The Kroz (wiki link for the kids) is 23 years old, but I wasn't there for the begining of that so I guess I'd put it around 20 years.
I don't think anyone knew what a frame rate was back then though.
Anyone remember the gem F-117 I need to track down a copy of that when I get home tonight.
I don't think anyone knew what a frame rate was back then though.
Anyone remember the gem F-117 I need to track down a copy of that when I get home tonight.
Well I'm 31 now in got into PC gaming when I was about 16. I didn't have a Windows machine to start out. I started with DOS 6.2. I had a crappy 386SX 16MHz to start and had to go up from there. My 486DX2 66 ran Doom and the other games of the day OK, but Descent brought it to it's knees. Forget about Quake or anything like that. Of course it didn't come out until I had a 486DX4 100.
Pisha... I started with a Trash-80 (TRS) and TI-994A... Then I moved up to my dad's 8086 and 8088. To play games back then I first had to write it out of code... Sometimes my own invention, sometimes computing magazines would have some in the back... Back before hard-drives and all the fun conveniences of today. I had to save/load stuff from tape... IF event that... which was the heavenly step up from punch cards...
Still, you're just a baby to 'gaming' when you look at the big picture. Even still, am I... No offense obviously. And I've been doing it for over 30 years...
I've only been PC gaming since the modern era. Console gaming all my life. In any case punch cards and tape backups have no real relevance to modern gaming. I stand by my statement that 20-25FPS wasn't stellar performance. However multiplayer gaming wasn't the big deal it is now, and this was sufficient for the single player experience, and serial / LAN gaming of the day.
How can you say that. It was the only way we had to keep our software from back then without coding them all over again. It's all we had. It led to the floppy drives and hard drives of today. How can you say they have no relevance to modern gaming?!? They are forms of data storage... One way or the other they led to modern gaming. Even those POS consoles (sorry, can't stand em, they ruin good gaming) have HDDs now-a-days...
And you think the 386 defines the modern era? No.. That started with the 8086...
I agree with that (never denied it). Personally I can't stand < 60 FPS myself. I get headaches too easily with FPS below that.
Good lord. I didn't take a piss in your breakfast cereal. I didn't define the modern era in any specific way. Though now that you mention it I'd say that started with the original XT and AT systems. Both of which I've worked on. My first jobs in this industry had me working on archaic hardware. So while I didn't experience this stuff when it was new, I did experience some of it. Though nothing going back to the punch card days. In any case, it may have started with punch cards and tapes, but the fact is that has little relevance today. Few gamers know what those items are and if they do, they know nothing outside of historical context. It has as much to do with gaming as the VAX machines or pocket calculators. Outside of historical context, it's meaningless. These items aren't referenced every day and aren't easy to relate to. They are too different from what we use today. So different they are archaic and almost alien. Beyond the philosophical argument that we wouldn't be at "X" without "A" coming first, how exactly is it relevant to the modern gamer? How does being aware of such things impact your technical ability to troubleshoot a system, build modern systems, install software or play games? Beyond relating legends of such devices to your customers, coworkers and friends, such knowledge serves no practical purpose. It has as much to do with modern gaming as a Roman chariot has to do with a 2011 Mustang GT.
That doesn't mean that such knowledge is totally useless or that we shouldn't bother knowing this history, but to claim relevance beyond anecdotes used in conversation seems silly to me. This knowledge does nothing to further your understanding of what we use today. Anymore than the abacus helps one understand the technical workings of the pocket calculator or the GPS system in your car.
Last edited:
You've proved a theory that 'modern' gamers just have no understanding of what is/were important and how things of old developed to what you love today.
What you said basically is the same as saying the old Ford Model T has nothing to do with a Modern Mustang... Or even further how other inventions like the steam tractor by Cugnot in the 1700's led to the original Ford. It's all one idea leading to a better one. Saying it has no relevance to modern gaming is just ignorant. You fail to see how one is the cause of the other, it's completely relevant. It may not be OBVIOUS, but it is indeed relevant. Historical context is entirely relevant. You wouldn't have one without the other.
You wouldn't have your modern SSDs without the punch-cards and tape drives of old. You wouldn't have your modern graphics cards without the old 16 color generators of old. You wouldn't even have your new LCDs without the old Tubes of old... You don't get one without the other. Therefore, quite simply they are completely relevant accept to the very narrow minded.
OMG; then you edit it to just prove your narrow-mindedness... You don't get one without the other. The Roman Chariot is completely relevant to the existence of the modern Mustang. It actually amazes me you can even think otherwise... I truly did LOL. Thanks for that.
No I didn't. The punch card systems of old are further removed from the modern computer than the Model T is from the modern Mustang GT or any modern car. Again I see how one is the cause of the other, but again beyond historically based conversational anecdotes, how does such knowledge help the modern gamer or even a modern computer technician? I know what these items are and this knowledge doesn't aide me in any way. It doesn't make my job any easier outside of being able to grasp the basics of stories some old timers relate to me. It is mildly interesting and little more.
It's obvious we wouldn't have the modern without the archaic. Specifically the basic wheel. I know this. You know this, most people know this. Tell me, what does that knowledge really do for me? How does it help me enrich my life?
You know, in thinking about it, I can say, yeah I can apply the basic principals of the Roman chariot to modern vehicles in the most fundemental way, and use it as the basis for more complex knowledge. Still I don't think that it would help me change a tire, or rebuild a differential, or replace the heads on an LS2 engine. Perhaps this wasn't the best example. When I learned computers, I learned about punch cards after the fact and I knew how hard drives worked prior to gaining that knowledge. I suppose this is why I find it largely irrelevant. I never related to it because my understanding of computer hardware wasn't based on that stepping stone first.
Last edited:
Great memories. I had a 386SX-16 that was overclocked to 20 by replacing the clock gen on the motherboard. Those were the good old days of overclocking. I never really thought about it before, but I guess the 486DX2 and DX4 were the first real chips with multipliers.